Figure 5.

The rate of change in the place cell population mirrors the rate of change in the visual scene

Trajectories were split into 1-cm intervals, and the Euclidean distance between adjacent activity vectors was calculated for the largest environment (see STAR Methods on population activity change).

(A and B) For trajectories orthogonal to the wall (A), the rate of change in the place cell population was greater when animals were close to the wall—an effect that is not observed for trajectories running parallel to the wall (B). Pairwise post hoc tests adjusted for multiple comparisons using Benjamini-Hochberg (non-negative) correction. The cartoon above each plot indicates the wall distances and includes locations in the environment in green bands, and the green sectors on the circles indicate the sampled movement directions. The legend above indicates spatial bins (left) and movement directions (right) used to plot data. The same observations were made in the other large environments (Figures S5A and S5B). ∗p = 0.05; ns, not significant.

(C) An example 8-m trajectory in the largest environment from animal 5.

(D) Time series data for the trajectory in (C) showing a tight coupling between Z scored rate of change in the hippocampal population and reconstructed visual scene (Figure S5D).

(E and F) Spatially averaged rates of change for place cell population and visual scenes (data from all rats, no smoothing) are similar, both being accentuated toward walls, corners, and cues. Note the local increase at cue boundaries observed for both visual and population change (black lines in C indicate location of wall-mounted cues).

(G) Population activity change correlates more strongly (bars indicate correlation across all animals, points indicate correlations per animal) with visual change (r = 0.60) than speed (r = −0.30), turning rate (r = 0.27), distance to the nearest walls orthogonal (r = −0.17) or parallel (r = 0.05) to the rats’ motion, and path integration since the rat last touched a wall (r = −0.06).

(H) Filtering by heading direction reveals how the change in both the visual scene and population activity depends on proximity to walls and their orientation relative to the direction of travel.